谢谢
还真是 numpy 版本的问题
之前是 1.17.4，升级到最新版 (1.18.1)，再执行 classify_with_model.py 就通过了。

2020-03-10 17:59:11.842 | INFO     | stagesepx.reporter:draw:261 - save report to 2020031017591156.html

按work_with_stagesepx实践了一下 auto、manual、dynamic
其中，auto、manual 成功，尝试 dynamic 第三步执行 classify_with_model.py 分类时总是报错，产生的中间数据不对，KerasClassifier 分析原理的具体细节不了解，请楼主指点一下，谢谢

2020-03-10 15:38:40.925 | DEBUG    | stagesepx.hook:do:24 - execute hook: GreyHook, frame id: 1
Traceback (most recent call last):
  File "classify_with_model.py", line 25, in <module>
    classify_result = cl.classify(video, stable, keep_data=True)
  File "/usr/local/lib/python3.7/site-packages/stagesepx/classifier/base.py", line 359, in classify
    result = self._classify_frame(frame, *args, **kwargs)
  File "/usr/local/lib/python3.7/site-packages/stagesepx/classifier/keras.py", line 198, in _classify_frame
    return self.predict_with_object(frame.data)
  File "/usr/local/lib/python3.7/site-packages/stagesepx/classifier/keras.py", line 193, in predict_with_object
    frame = np.expand_dims(frame, axis=[0, -1])
  File "<__array_function__ internals>", line 6, in expand_dims
  File "/Users/aa/Library/Python/3.7/lib/python/site-packages/numpy/lib/shape_base.py", line 577, in expand_dims
    if axis > a.ndim or axis < -a.ndim - 1:
TypeError: '>' not supported between instances of 'list' and 'int'